3 Review(s)

This monograph opens up new horizons for engineers and researchers inacademia and in industry dealing with or interested in new developments in thefield of system identification and control. It emphasizes guidelines forworking solutions and practical advice for their implementation rather than thetheoretical background of Gaussian process (GP) models. The book demonstratesthe potential of this recent development in probabilistic machine-learningmethods and gives the reader an intuitive understanding of the topic. Thecurrent state of the art is treated along with possible future directions forresearch. Systems control design relies on mathematical models and these may bedeveloped from measurement data. This process of system identification, whenbased on GP models, can play an integral part of control design in data-basedcontrol and its description as such is an essential aspect of the text. Thebackground of GP regression is introduced first with system identification andincorporation of prior knowledge then leading into full-blown control. The bookis illustrated by extensive use of examples, line drawings, and graphicalpresentation of computer-simulation results and plant measurements. Theresearch results presented are applied in real-life case studies drawn fromsuccessful applications including: a gas-liquid separator control; urban-traffic signal modelling and reconstruction; and prediction of atmospheric ozone concentration. A MATLAB (R) toolbox, for identification and simulation ofdynamic GP models is provided for download.